Large-scale Software-Defined Networking Research Articles

Machine-Learning-Based Traffic Classification in Software-Defined Networks

Many research efforts have gone into upgrading antiquated communication network infrastructures with better ones to support contemporary services and applications. Smart networks can adapt to new technologies and traffic trends on their own. Software-defined networking (SDN) separates the control plane from the data plane and runs programs in one place, changing network management. New technologies like SDN and machine learning (ML) could improve network performance and QoS. This paper presents a comprehensive research study on integrating SDN with ML to improve network performance and quality-of-service (QoS). The study primarily investigates ML classification methods, highlighting their significance in the context of traffic classification (TC). Additionally, traditional methods are discussed to clarify the ML outperformance observed throughout our investigation, underscoring the superiority of ML algorithms in SDN TC. The study describes how labeled traffic data can be used to train ML models for appropriately classifying SDN TC flows. It examines the pros and downsides of dynamic and adaptive TC using ML algorithms. The research also examines how ML may improve SDN security. It explores using ML for anomaly detection, intrusion detection, and attack mitigation in SDN networks, stressing the proactive threat-detection and response benefits. Finally, we discuss the SDN-ML QoS integration problems and research gaps. Furthermore, scalability and performance issues in large-scale SDN implementations are identified as potential issues and areas for additional research.

A Scalable Monitoring System for Software Defined Networks

Monitoring functionality is an essential element of any network system. Traditional monitoring solutions are mostly used for manual and infrequent network management tasks. Software-defined networks (SDN) have emerged with enabled automatic and frequent network reconfigurations. In this paper, a scalable monitoring system for SDN is introduced. The proposed system monitors small, medium, and large-scale SDN. Multiple instances of the proposed monitoring system can run in parallel for monitoring many SDN slices. The introduced monitoring system receives requests from network management applications, collects considerable amounts of measurement data, processes them, and returns the resulting knowledge to the network management applications. The proposed monitoring system slices the network (switches and links) into multiple slices. The introduced monitoring system concurrently monitors applications for various tenants, with each tenant's application running on a dedicated network slice. Each slice is monitored by a separate copy of the proposed monitoring system. These copies operate in parallel and are synchronized. The scalability of the monitoring system is achieved by enhancing the performance of SDN. In this context, scalability is addressed by increasing the number of tenant applications and expanding the size of the physical network without compromising SDN performance.

A Novel Mechanism for Detection of Address Resolution Protocol Spoofing Attacks in Large-Scale Software-Defined Networks

A Novel Mechanism for Detection of Address Resolution Protocol Spoofing Attacks in Large-Scale Software-Defined Networks

Feature-based real-time distributed denial of service detection in SDN using machine learning and Spark

Recently, software defined networking (SDN) has been deployed extensively in diverse practical domains, providing a new direction in network management by separating the control plane from the data plane. Nevertheless, SDN is vulnerable to distributed denial of service (DDoS) attacks resulting from its centralized controller. Several studies have been suggested to address the DDoS attacks in SDN utilizing machine learning approaches. However, these approaches are resource-intensive and cause performance degradation since they cannot perform effectively in large-scale SDN networks that generate vast traffic statistics. To handle all these challenges, we build a DDoS attack detection model in SDN using Spark as a big data tool to overcome the limitations of conventional data processing methods. Four machine learning algorithms are employed. The decision tree (DT) is elected to be used for real-time deployment based on the performance results, which indicates that it has the best accuracy of 0.936. The model performance is compared with state-of-the-art and shows an overall better performance.

Feature-based real-time distributed denial of service detection in SDN using machine learning and Spark

Recently, software defined networking (SDN) has been deployed extensively in diverse practical domains, providing a new direction in network management by separating the control plane from the data plane. Nevertheless, SDN is vulnerable to distributed denial of service (DDoS) attacks resulting from its centralized controller. Several studies have been suggested to address the DDoS attacks in SDN utilizing machine learning approaches. However, these approaches are resource-intensive and cause performance degradation since they cannot perform effectively in large-scale SDN networks that generate vast traffic statistics. To handle all these challenges, we build a DDoS attack detection model in SDN using Spark as a big data tool to overcome the limitations of conventional data processing methods. Four machine learning algorithms are employed. The decision tree (DT) is elected to be used for real-time deployment based on the performance results, which indicates that it has the best accuracy of 0.936. The model performance is compared with state-of-the-art and shows an overall better performance.

A Multi-Controller Placement Strategy for Hierarchical Management of Software-Defined Networking

Software-Defined Networking (SDN) is a new architecture with symmetric/asymmetric network structures that separates the control plane of network devices from the data plane, and a Controller Placement Problem (CPP) is a critical management problem in SDN. The main research content of the CPP is to determine the number and location of controllers placed in a network topology, as well as the connection relationship between controllers and switches. However, traditional CPP solutions based on symmetric/asymmetric structures may not be efficient to meet the increasing requirements of SDN applications. In order to improve the CPP solutions from the viewpoint of hierarchical management, this paper considers the CPP solutions as a multi-objective optimization problem based on symmetric/asymmetric structures in the SDN architecture. Thus, this paper then proposes a multi-controller placement strategy based on an improved Harris Hawks Optimization algorithm. Firstly, the local controller load is limited, and a Sin chaotic map is introduced to initialize the CPP scheme. The total latency of the network, the reliability of the node, the total failure rate of the link and the total placement cost are seriously considered when placing the controllers. Secondly, a Cos nonlinear function is added to the global search. A dynamic adaptive weight factor is used to smooth the switching approach between the global search and the local search, so as to enhance the global search ability. Then, a Cauchy variation perturbation is added to the obtained CPP scheme to strengthen the diversity of CPP schemes, and the CPP scheme with the Pareto front is finally solved. The topology simulation of three real large-scale SDN networks shows that the proposed strategy, based on an improved Harris Hawks Optimization algorithm, has more robust advantages in comparison to other algorithms.

ProbD: Faulty Path Detection Based on Probability in Software-Defined Networking

With the increasing number of switches in Software-Defined Networking (SDN), there are more and more faults rising in the data plane. However, due to the existence of link redundancy and multi-path forwarding mechanisms, these problems cannot be detected in time. The current faulty path detection mechanisms have problems such as the large scale of detection and low efficiency, which is difficult to meet the requirements of efficient faulty path detection in large-scale SDN. Concerning this issue, we propose an efficient network path fault testing model ProbD based on probability detection. This model achieves a high probability of detecting arbitrary path fault in the form of small-scale random sampling. Under a certain path fault rate, ProbD obtains the curve of sample size and probability of detecting arbitrary path fault by randomly sampling network paths several times. After a small number of experiments, the ProbD model can correctly estimate the path fault rate of the network and calculate the total number of paths that need to be detected according to the different probability of detecting arbitrary path fault and the path fault rate of the network. The final experimental results show that, compared with the full path coverage test, the ProbD model based on probability detection can achieve efficient network testing with less overhead. Besides, the larger the network scale is, the more overhead will be saved.

Multi-Controller Placement Optimization Using Naked Mole-Rat Algorithm over Software-Defined Networking Environment

Software Defined Networking (SDN) is the novel networking paradigm where decoupling of the control plane from the data plane has its inherent advantages. Controller Placement Problem (CPP) involves placing the optimal number of controllers at the appropriate locations while meeting prerequisites such as latency, load balancing, energy and computational time. To achieve scalability, deployment of multiple controllers on large-scale SDN is one of the key challenges. CPP can be addressed as a multi-objective combinatorial optimization problem whose solution is a trade-off between multiple optimization parameters. In this paper, a novel population-based meta-heuristic algorithm viz. Naked Mole-Rat (NMR) Algorithm has been proposed to optimize the location for controller placement based on Switch-Controller (SC), Controller-Controller (CC) latency while maintaining load balancing among the controllers. The ideas and mechanisms are illustrated using two publicly available standard topologies viz. Ernet and Savvis. The controller localization approach implemented with NMR algorithm has slightly a better result as compared with the Bat algorithm.

A Novel Strategy for Computing Routing Paths for Software-Defined Networks Based on MOCell Optimization

Software-defined networking (SDN) is the fastest growing and most widely deployed network infrastructure due to its adaptability to new networking technologies and intelligent applications. SDN simplifies network management and control by separating the control plane from the data plane. The SDN controller performs the routing process using the traditional shortest path approach to obtain end-to-end paths. This process usually does not consider the nodes’ capacity and may cause network congestion and delays, affecting flow performance. Therefore, we evaluate the most conventional routing criteria in the SDN scenario based on Dijkstra’s algorithm and compare the found paths with our proposal based on a cellular genetic algorithm for multi-objective optimization (MOCell). We compare our proposal with another multi-objective evolutionary algorithm based on decomposition (MOEA/D) for benchmark purposes. We evaluate various network parameters such as bandwidth, delay, and packet loss to find the optimal end-to-end path. We consider a large-scale inter-domain SDN scenario. The simulation results show that our proposed method can improve the performance of data streams with TCP traffic by up to 54% over the traditional routing method of the shortest path and by 33% for the highest bandwidth path. When transmitting a constant data stream using the UDP protocol, the throughput of the MOCell method is more than 1.65% and 9.77% for the respective paths.

Dynamic Routing Optimization Algorithm for Software Defined Networking

Time and space complexity is the most critical problem of the current routing optimization algorithms for Software Defined Networking (SDN). To overcome this complexity, researchers use meta-heuristic techniques inside the routing optimization algorithms in the OpenFlow (OF) based large scale SDNs. This paper proposes a hybrid meta-heuristic algorithm to optimize the dynamic routing problem for the large scale SDNs. Due to the dynamic nature of SDNs, the proposed algorithm uses a mutation operator to overcome the memory-based problem of the ant colony algorithm. Besides, it uses the box-covering method and the k-means clustering method to divide the SDN network to overcome the problem of time and space complexity. The results of the proposed algorithm compared with the results of other similar algorithms and it shows that the proposed algorithm can handle the dynamic network changing, reduce the network congestion, the delay and running times and the packet loss rates.

Flow aggregation for large-scale SDNs with scattered address space allocation

Software-Defined Networking (SDN) has obtained a lot of attention in the last decade and has played a significant role in the development of next-generation networks (NGN). IP networks can also benefit from the SDN evolution to fulfill the data traffic booming. However, the transition of the traditional networking model to SDN architectures poses scalability issues due to the possible flow entry explosion in SDN switches. The limited size of flow-table of SDN switches is not sufficient to handle thousands upon thousands of flows in a large-scale IP network. On the other hand, the interleaved allocation of non-contiguous IP addresses also leads to inefficient routing aggregation and reduces the feasibility of the serious implementation of SDN severely. Therefore, we propose an aggressive flow aggregation scheme—Destination Address Translation and Source-Port Translation on Demand (DATSPToD), which is based on the modified address and port rewriting. DATSPToD enables the aggregation of flow entries in SDNs by translating the destination addresses of multiple same-destination flows with different-source into one flow entry, thus significantly reducing the volume of flow-table occupancy of core-layer SDN switches, even in freely scattered IP address space environments. Simulation results show that DATSPToD outperforms non-aggregation and both wildcard aggregation schemes for a significant reduction of the flow-table occupancy under varied traffic patterns and topologies, especially in large-scale SDNs such as the Internet during the SDN migration period. Compared to the competed schemes under varied traffic patterns and topologies, the simulation results of the DATSPToD could significantly reduce flow-table occupancies by at least1 13%, and by up to 52%. The table-miss rates could be reduced from 11% to 66%.

MaxHadoop: An Efficient Scalable Emulation Tool to Test SDN Protocols in Emulated Hadoop Environments

This paper presents MaxHadoop, a flexible and scalable emulation tool, which allows the efficient and accurate emulation of Hadoop environments over Software Defined Networks (SDNs). Hadoop has been designed to manage endless data-streams over networks, making it a tailored candidate to support the new class of network services belonging to Big Data. The development of Hadoop is contemporary with the evolution of networks towards the new architectures “Software Defined.” To create our emulation environment, tailored to SDNs, we employ MaxiNet, given its capability of emulating large-scale SDNs. We make it possible to emulate realistic Hadoop scenarios on large-scale SDNs using low-cost commodity hardware, by resolving a few key limitations of MaxiNet through appropriate configuration settings. We validate the MaxHadoop emulator by executing two benchmarks, namely WordCount and TeraSort, to evaluate a set of Key Performance Indicators. The tests’ outcomes evidence that MaxHadoop outperforms other existing emulation tools running over commodity hardware. Finally, we show the potentiality of MaxHadoop by utilizing it to perform a comparison of SDN-based network protocols.

Path optimization of box-covering based routing to minimize average packet delay in software defined network

A routing algorithm plays a major role in Data communication across an inter-network. Software-Defined Networking (SDN) is a new era of computer networking that separates Data plane from Control plane which controls, changes and manages the network behavior dynamically via open interfaces with the help of SDN Controllers. In SDN, the networking devices route packets in the form of flows with the help of flow rules given by the controllers and the flow rules are stored in the corresponding flow tables of the networking devices. The Box-Covering (BC) algorithm is an existing algorithm used in SDN for achieving renormalization of networks by calculating the fractal dimension of networks and covering the network with the minimum possible number of boxes and Dijkstra’s algorithm is used for calculating shortest paths. The proposed work focuses on Path Optimization in the Box-Covering-Based Routing (BCR) algorithm, which is an existing algorithm used in large-scale SDN. In the proposed work, a Link-Weight system has been used for bounding the Link Utilization of the shortest path between the Source and the Destination to minimize the Average Packet Delay in the Network. The results show that the proposed work reduces the Average Packet Delay compared to the existing algorithms such as Dijkstra’s algorithm, Graph Compression algorithm and BCR algorithm and also improves the performance of the network.

A DAG-Based Forwarding Paradigm for Large Scale Software Defined Networks

The Software Defined Network (SDN) paradigm represents a major breakthrough in the networking field, due to its unprecedented capabilities in terms of flexibility and programmability. SDNs have been successfully deployed in data centers and small to medium enterprises. However, adopting the SDN paradigm in the context of wide area networks is more challenging, due to a number of factors including the higher probability that node and link failures occur and the unavailability of a dedicated control channel. In this paper, we present a DAG (Directed Acyclic Graph) based forwarding paradigm addressing the challenges that arise when the SDN concept is applied to large scale networks. Specifically, the proposed paradigm aims to limit the number of entries required on the SDN switches, to provide a fast local restoration of single node/link failures without the intervention of the SDN controller and to prevent the possibility of having inconsistent forwarding tables during updates. The proposed paradigm does not require any extension to the OpenFlow protocol and we show how it can be implemented by only using standard features. The DAG-based forwarding paradigm requires to compute a DAG between every pair of ingress-egress switches and to design an index-based hashing scheme to balance the load across the paths in the DAG while avoiding TCP reordering issues. In this paper, we present heuristic algorithms providing a solution to such problems and report the results of a simulation study conducted to assess the performance of the proposed forwarding paradigm.

A Hierarchical Distributed Control Plane for Path Computation Scalability in Large Scale Software-Defined Networks

Given the shortcomings of traditional networks, software-defined networking (SDN) is considered as the best solution to deal with the constant growth of mobile data traffic. SDN separates the data plane from the control plane, enabling network scalability, and programmability. Initial SDN deployments promoted a centralized architecture with a single controller managing the entire network. This design has proven to be unsuited for nowadays large-scale networks. Though multi-controller architectures are becoming more popular, they bring new concerns. One critical challenge is how to efficiently perform path computation in large networks considering the substantial computational resources needed. This paper proposes HiDCoP, a distributed high-performance control plane for path computation in large-scale SDNs along with its related solutions. HiDCoP employs a hierarchical structure to distribute the load of path computation among different controllers, reducing therefore the transmission overhead. In addition, it uses node parallelism to accelerate the performance of path computation without generating high control overhead. Simulation results show that HiDCoP outperforms existing schemes in terms of path computation time, end-to-end delay, and transmission overhead.

Efficient Recovery Path Computation for Fast Reroute in Large-Scale Software-Defined Networks

With an increasing demand for resilience in software-defined networks (SDN), it becomes critical to minimize service recovery delay upon route failures. Fast reroute (FRR) mechanisms are widely used in IP and MPLS networks by computing the recovery path before a failure occurs. The centralized control plane in SDN can potentially enhance path computation, so that FRR path computation can better scale in SDN than in traditional networks. However, the traditional FRR path computation algorithms could lead to a poor performance in large-scale SDN. The problem can become more severe for a highly dynamic network, which often sees dozens of failures or configuration changes in any single day. We propose a new algorithm that exploits pruned searching to quickly compute recovery paths for all-pair switches/hosts upon a link failure. For applications requiring stringent path robustness levels, we also extend this algorithm to quickly find the shortest guaranteed-cost path, which ensures that the recovery path used upon on-path link failures has the minimum cost. Compared with traditional solutions, our evaluations show that our algorithm is about 8 ~ 81 times faster than the practical implementation, 1.93 ~ 3.11 times faster than the state-of-the-art solution. Our results also show that the shortest guaranteed-cost path can reduce the cost of the recovery path significantly. Moreover, we design a prototype to show how to deploy our algorithm in an OpenFlow network.

A new framework for reliable control placement in software-defined networks based on multi-criteria clustering approach

In large-scale software-defined networks (SDNs), multiple controllers are deployed. Each controller has a logically centralized vision of the network that is used to manage a set of switches. In SDN, a challenge known as controller placement problem arises which is very important to specify the number of controllers that are needed and where they should be deployed. These specifications affect the performance of the network. Meanwhile, the assignment of switches to the controllers plays key role in the quality of solution in this problem. However, recent studies focus more on simply assigning switches to their closest controllers based on propagation delay between controller and switches. In this paper, a new controller placement framework is designed which considers both control plane architecture and relation between control and data planes. This framework is considered as a multi-objective optimization model with two objective functions to minimize the flow setup time and inter-controller latency. Furthermore, we propose a new model for flow setup time function that considers all affected metrics in the placement. To solve the framework, a multi-objective algorithm called non-dominated sorting moth flame controller placement optimizer is designed. To this end, we adapt the best–worst multi-criteria decision-making method considering three metrics, namely hop count, propagation latency and link utilization, to assign switches to controllers. A heuristic approach is also used to assign a path between switch and its controller using above three metrics and path reliability. We run theses three models iteratively to find the best location for controllers, the best switch assignment to controller and also find the best route in the network. We compare our proposed framework with other models using expected path loss and link load balancing metrics. Our performance evaluations on real wide area network topologies show the efficiency of the proposed framework.

Dynamic slave controller assignment for enhancing control plane robustness in software-defined networks

Multi-controller is a scalable control plane solution for the large-scale Software-Defined Networking (SDN). To achieve high resilience, an SDN switch can connect one master controller for normal operation and one slave controller that backup the function of the master controller. Once the master controller fails, one of the slave controllers will be assigned to switches to works as the new master controller. However, the inappropriate slave controller assignment may cause controller chain failure, where running out of the capacity of the assigned controller, even crash the entire network. In this paper, we propose a dynamic slave controller assignment that prevents the network crash by planning slave controller assignment ahead of the controller failures. We first describe the controller chain failure phenomenon: due to unreasonable slave controller assignment, the entire network may crash when one controller fails. To prevent the phenomenon, we formulate the slave controller assignment problem as a multi-objective mixed optimization problem that considers latency, load balancing and robustness, and prove its NP-complete complexity. We solve the problem with a dynamic slave controller assignment (DSCA) scheme. It firstly checks whether there are controller failures in state detection module, then completes the elastic slave assignment and generates a new slave assignment for switches in efficient slave assignment module. Finally, in role adjustment module, it changes the roles of some controllers and reconnects switches. Simulation results show our solution can decrease the worst case latency under controller failures by 35.1% averagely, and reduce the probability of network crash.

Chaotic salp swarm algorithm for SDN multi-controller networks

Software-defined networking (SDN) is a novel network paradigm that enables flexible management for networks. However, with the increase in network capacity, a single controller of SDN has many limitations on both performance and scalability. Distributed multi-controller deployment is a promising method to satisfy fault tolerant and scalability. There are still open research issues related to controllers placement, and the optimal number of deployed controllers. In this paper, a dynamic optimization algorithm that is based on the Salp Swarm Optimization Algorithm (SSOA) is developed with the introduction of chaotic maps for enhancing the optimizer’s performance. The algorithm dynamically evaluates the optimum number of controllers and the optimal connections between switches and controllers in large scale SDN networks. In order to evaluate the proposed algorithm, several experiments were conducted and implemented in various scenarios. Moreover, the algorithm was compared to the linear and meta-heuristic algorithms. Simulation results show that the proposed algorithm outperforms meta-heuristic algorithms and a game theory based algorithm in terms of execution time and reliability.

Joint Optimization of Flow Table and Group Table for Default Paths in SDNs

Software-defined networking (SDN) separates the control plane from the data plane to ease network management and provide flexibility in packet routing. The control plane interacts with the data plane through an interface that configures the forwarding tables, usually including a flow table and a group table, at each switch. Due to high cost and power consumption of ternary content addressable memory, commodity switches can only support flow/group tables of limited size, which presents serious challenge for SDN to scale to large networks. One promising approach to address the scalability problem is to deploy aggregate default paths specified by wildcard forwarding rules. However, the multi-dimensional interaction among numerous system parameters and performance/scalability considerations makes the problem of setting up the flow/group tables at all switches for optimal overall layout of default paths very challenging. This paper studies the joint optimization of flow/group tables in the complex setting of large-scale SDNs. We formulate this problem as an integer linear program, and prove its NP-hardness. An efficient algorithm with bounded approximation factors is proposed to solve the problem. The properties of our algorithm are formally analyzed. We implement the proposed algorithm on an SDN test bed for experimental studies and use simulations for large-scale investigation. The experimental results and simulation results show that, under the same number of flow entries, our method can achieve better network performance than the equal cost multipath while reducing the use of group entries by about 74%. Besides, our method can reduce the link load ratio and the number of flow entries by approximately 13% and 60% compared with DevoFlow with 10% additional group entries.